> ## Documentation Index
> Fetch the complete documentation index at: https://learn.nexudus.com/llms.txt
> Use this file to discover all available pages before exploring further.

# OpenAI

> Use your OpenAI account to answer help-desk messages from your FAQs and draft customers' professional profiles.

## Overview

The **OpenAI (beta)** integration uses your own OpenAI account to:

* **Answer help-desk messages** using your published [FAQs](https://learn.nexudus.com/platform/content/faq.md).
* **Draft professional profiles** for customers, or let customers draft their own.

The integration is free in Nexudus. OpenAI charges your account for each request, by the number of tokens used — set a monthly limit in your OpenAI billing settings to control costs. When the limit is reached, automatic answers stop.

> [!NOTE]
>
> For AI that works across your inbox, bookings and more, see [AI agents](https://learn.nexudus.com/platform/ai/overview.md).

## Connecting OpenAI

1. In your OpenAI account, add a payment method and create a secret API key. Copy it — you can't see it again.
2. Go to **Settings → Integrations → OpenAI (beta)**, turn on **Enable OpenAI integration.**, and enter the **API Key**.
3. Choose the options below, and save.

![OpenAI settings](https://mintcdn.com/nexudus/_vKYV6BblXHm7HHO/images/integrations/int-openai.png?fit=max\&auto=format\&n=_vKYV6BblXHm7HHO\&q=85\&s=ca868b8bbc5cad45fdbd40ba87ba4a57)

| Setting | What it does |
| - | - |
| **Respond to help-desk messages using OpenAI conversational models based on your FAQ articles.** | Replies to help-desk messages using your FAQs. See [Help-desk](https://learn.nexudus.com/platform/operations/help-desk.md). |
| **Let customers draft their professional profiles using OpenAI** | Customers can generate a profile draft. |
| **Draft a professional profile for every new customer (unpublished by default).** | Creates a draft profile for each new customer. |

## Getting good answers

Only **published** FAQs are used. Write them in short, simple sentences, cover one topic per FAQ — for example separate Wi-Fi, printing and hardware FAQs — and use the same wording your customers do. Test with real questions from closed help-desk messages before relying on it.

## Related

* [FAQs](https://learn.nexudus.com/platform/content/faq.md) — the answers OpenAI draws on
* [Help-desk](https://learn.nexudus.com/platform/operations/help-desk.md) — help-desk messages


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.